UM  > INSTITUTE OF MICROELECTRONICS
Affiliated with RCfalse
Status已發表Published
A 108-nW 0.8-mm 2 Analog Voice Activity Detector Featuring a Time-Domain CNN With Sparsity-Aware Computation and Sparsified Quantization in 28-nm CMOS
Chen, Feifei; Un, Ka Fai; Yu, Wei Han; Mak, Pui In; Martins, Rui P.
2022-07
Source PublicationIEEE Journal of Solid-State Circuits
ISSN0018-9200
Volume57Issue:11Pages:1-10
Abstract

This article reports a passive analog feature extractor for realizing an area-and-power-efficient voice activity detector (VAD) for voice-control edge devices. It features a switched-capacitor circuit as the time-domain convolutional neural network (TD-CNN) that extracts the 1-bit features for the subsequent binarized neural network (BNN) classifier. TD-CNN also allows area savings and low latency by evaluating the features temporally. The applied sparsity-aware computation (SAC) and sparsified quantization (SQ) aid in enlarging the output swing and reducing the model size without sacrificing the classification accuracy. With these techniques, the diversified output also aids in desensitizing the 1-bit quantizer from the offset and noise. The TD-CNN and BNN are trained as a single network to improve the VAD reconfigurability. Benchmarking with the prior art, our VAD in 28-nm CMOS scores a 90% (94%) speech (non-speech) hit rate on the TIMIT dataset with small power (108 nW) and area (0.8 mm 2) . We can configure the TD-CNN as a feature extractor for keyword spotting (KWS). It achieves a 93.5% KWS accuracy with the Google speech command dataset (two keywords). With two TD-CNNs operating simultaneously to extract more features, the KWS accuracy is 94.3%.

KeywordApproximate Computing Convolutional Neural Network (Cnn) Feature Extraction Keyword Spotting (Kws) Quantization Reconfigurable Sparsity Switched-capacitor Circuits Voice Activity Detection (Vad)
DOI10.1109/JSSC.2022.3191008
URLView the original
Indexed BySCIE
Language英語English
WOS Research AreaEngineering
WOS SubjectEngineering, Electrical & Electronic
WOS IDWOS:000829074600001
Fulltext Access
Citation statistics
Cited Times [WOS]:0   [WOS Record]     [Related Records in WOS]
Document TypeJournal article
CollectionINSTITUTE OF MICROELECTRONICS
Faculty of Science and Technology
DEPARTMENT OF ELECTRICAL AND COMPUTER ENGINEERING
Corresponding AuthorUn, Ka Fai
AffiliationUniversity of Macau
First Author AffilicationUniversity of Macau
Corresponding Author AffilicationUniversity of Macau
Recommended Citation
GB/T 7714
Chen, Feifei,Un, Ka Fai,Yu, Wei Han,et al. A 108-nW 0.8-mm 2 Analog Voice Activity Detector Featuring a Time-Domain CNN With Sparsity-Aware Computation and Sparsified Quantization in 28-nm CMOS[J]. IEEE Journal of Solid-State Circuits,2022,57(11):1-10.
APA Chen, Feifei,Un, Ka Fai,Yu, Wei Han,Mak, Pui In,&Martins, Rui P..(2022).A 108-nW 0.8-mm 2 Analog Voice Activity Detector Featuring a Time-Domain CNN With Sparsity-Aware Computation and Sparsified Quantization in 28-nm CMOS.IEEE Journal of Solid-State Circuits,57(11),1-10.
MLA Chen, Feifei,et al."A 108-nW 0.8-mm 2 Analog Voice Activity Detector Featuring a Time-Domain CNN With Sparsity-Aware Computation and Sparsified Quantization in 28-nm CMOS".IEEE Journal of Solid-State Circuits 57.11(2022):1-10.
Files in This Item: Download All
File Name/Size Publications Version Access License
FINAL VERSION.PDF(1724KB)期刊论文作者接受稿开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Chen, Feifei]'s Articles
[Un, Ka Fai]'s Articles
[Yu, Wei Han]'s Articles
Baidu academic
Similar articles in Baidu academic
[Chen, Feifei]'s Articles
[Un, Ka Fai]'s Articles
[Yu, Wei Han]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Chen, Feifei]'s Articles
[Un, Ka Fai]'s Articles
[Yu, Wei Han]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: FINAL VERSION.PDF
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.